Innovation Management Watch Summary: “Breaking Barriers: Overcoming the Roadblocks to AI Adoption in New Product Development” by Dr. Robert G. Cooper

This week’s Innovation Management Watch Summary highlights Dr. Robert G. Cooper’s June 2024 article on why AI adoption in new product development (NPD) remains slower than expected despite strong early results. The article notes that early adopters such as GE, Siemens, Pfizer, and Nestlé report on major gains, including faster development, better design optimization, stronger launch planning, and increased innovation. Yet adoption remains limited: the paper cites a rise from 13% of firms using at least one AI application in NPD in early 2023 to 24% by early 2024, meaning most businesses still have not begun the journey. Dr. Cooper argues that the issue is not a lack of hype, but a set of practical and organizational barriers that can stop AI deployment before it scales.   

A central contribution of the article is its framing of seven main roadblocks to AI adoption in NPD. These include weak senior-management commitment, a poor or unproven business case, fear of failure, infrastructure and talent gaps, perceived risks, lack of trust, and ethical concerns. Some of these concerns are described as exaggerated or based on myths, but others are real and require active mitigation. The broader message is that companies should not treat AI adoption as a simple software purchase. It is a change effort that must be led, staged, and de-risked deliberately if it is to succeed.   

The article places particular weight on readiness and trust. According to the readiness findings summarized in the paper, the strongest factors linked to extensive AI use in NPD are visible performance improvements, leadership commitment to adopt, willingness to hand over some decision-making to AI, and the presence of a credible executive sponsor. One striking finding is that 98% of managers in the cited study were not prepared to hand over decision-making authority to AI, pointing to a major trust barrier. Dr. Cooper suggests that this hesitation reflects not only fear of the technology, but also limited AI literacy among managers and executives. Education is therefore presented as essential, both to build realistic understanding and to reduce myths around AI use.   

The business-case challenge is another major theme. The article argues that many firms still lack strong, fact-based evidence that AI improves NPD performance in ways that matter financially. At the same time, Dr. Cooper points to emerging evidence that some AI applications do show stronger links to performance than others. The chart on page 8 highlights the highest-impact uses as product testing, AI-assisted product design, rapid prototyping, and simulation models such as digital twins, with especially strong correlations to faster development and improved decision-making. By contrast, some earlier-stage uses that receive more attention, such as AI for ideation or building the business case itself, show less demonstrated impact in the current data. This leads to a practical recommendation: begin with lower-risk, easier-to-prove applications that can generate visible wins.   

Another important insight is that AI failure is often tied less to the technology itself than to poor business practices. The article reviews high reported failure rates for AI projects and argues that many breakdowns stem from avoidable issues such as weak change management, unclear user needs, poor-quality data, limited internal capabilities, and lack of a structured deployment roadmap. To reduce these risks, Dr. Cooper recommends a staged deployment process with gates, pilots, learning loops, and ongoing review rather than a one-shot rollout. The stage-gate diagram on page 10 reinforces this approach: start with front-end homework and a business case, move to acquisition or development, then pilots, scale-up, and post-implementation learning.   

The article concludes that AI adoption in NPD is a journey, not a single decision. Firms that begin with manageable applications, build trust through education and governance, and focus on visible business value are better positioned to scale over time. In that sense, the paper’s message is not simply that AI matters for innovation, but that disciplined adoption practices may matter just as much as the technology itself.   

This summary is based on Dr. Robert G. Cooper’s June 2024 article, “Breaking Barriers: Overcoming the Roadblocks to AI Adoption in New Product Development.” All rights to the original content remain with the respective copyright holders.